pantscat

@Elison, Sorry I probably wasn't clear. I meant what hardware do you use to send the SMS? do you plug a mobile phone into the server or use some sort of clever doohicky to do it?

@Elison - how do you do the SMS notification?

What server O/S are you running? You might want to think about running DFS and mirroring your data on the server at this smaller site - you can use site costs to ensure that the users at that site only access the data from that server (unless it's unavailable) plus it'll give you a level or redundancy as if your main file-server goes down users at the main site could access data across the 10mb link. Edit: Make it a DC and Global Catalog so that it can service the local user logons too... File access over a 10mb link isn't as slow as you'd think... I had a site with about 150 users and it worked just fine. Edit #2: I like HP Proliant DL servers... relatively small and rackable.

@sdc - That's very helpful. PM coming soon.

Bumpity bump. Anyone?

I've never had any issues with Sibelius licence server running on a Win2003 VM on an ESX 4.1 host. You'll probably be quite surprised once you get into as to how "affordable" server virtualisation can be (I was in a position whereby I was comparing it with replacing approximately 8 physical servers which were all rather old/stupidly specced/not-a-real-server). I'm running two hosts with a dedicated SAN and as the discussions with suppliers went on the prices came down further and further. Plus it helps if you can be flexible when you purchase - if you can buy when it's the hardware manufacturer's quarter end you'll probably be able to get some very good prices. I did with HP and NetApp...

@kennysarmy - Actually I was a bit wrong about WPAD - some devices wont support this out of the box (iPads, smart phones and the like). So... other people have been far more accurate by saying that you want a transparent proxy. Smoothwall does this rather nicely. The tricky part about a transparent proxy is identifying the user at the other end as it won't necessarily do any sort of authentication to identify the user - all you will see is that an anonymous user from ip address 1.2.3.4 is looking at Google, for example. It's up to you to decide whether this is acceptable. It may be possible to get your (hypothetical) wireless system to log all of the authentication attempts - e.g. record the date/time/IP address/username and then cross reference this against "naughty" behaviour on your transparent proxy. Essentially, unless you're allocating fixed IPs or IP reservations to client devices so that you know who is using which device (which doesn't help you with open access, multi-user PCs), then you'll have to give everyone the same level of internet access/filtering. I hope that makes sense. I should also point out that the transparent proxy would be set as the default gateway, and therefore the clients wouldn't have any proxy settings per se.

It is possible to have what you want... but this doesn't really correlate with your question "How does the internet work"... in my mind that's a completely different question. You want: -a wireless network system that uses Captive Portal authentication (my Ruckus system does this). - WPAD to push out your proxy settings Do you want to know more?

Hi all, I'm just starting to look at getting a good network monitoring system setup and one thing that I'd really like is to get SMS alerts. I've looked at Nagios, Paessler, Spiceworks and Foglight NMS and it appears that Foglight NMS will do everything I want it to do, and has enough "endpoint" licences (100) in the free edition to make it very good value for money. i.e. £0. Has anyone setup a similar system and used SMS alerts? What sort of hardware did you use? Ta, Ant

Hmm - I've not really touched much on multicast stuff. I know that the HP Switches have an option for turning on Multicast filtering (I think it looks at the destination IP/Mac addresses) and tries to send it to only those bits of the network that need to receive it. I must admit that I don't know a lot more than that!

@fracmo2000 - Aww, gee, shucks. Don't mention it.

Hi, I'm guessing it's trying to contact a licence server perhaps? First thing is to try and work out what port whatever the service running on the server is listening to. 1. Logon to the server 2. Go to a command prompt 3. Use "netstat -a -n -o" to list all the processes and ports 4. Hopefully you'll know or recognise the name of the executable and can match it to the port. Next step, once you've found out the port, is to reconfigure the HP switches in that area with an additional ACL to allow traffic from the client PCs to the server on that particular port. The best way to do this is to export the config to a text file - use something like "show run" from a telnet/SSH session and copy it to a text file - from here you'll be able to see the existing ACL statements and edit them or add a new one. Hope this helps, Ant

Perhaps just don't include the Fog client in the system build? or when installing the client don't select the "hostname changer" module.

Yep. But the IOS is very very different. Have you compared the two?

Just my 2p worth - when I was going through the motions of replacing my LAN last summer I had special bid pricing from HP Procurve, Linksys (bleurgh!), Cisco and Netgear. In the end, price-wise, it was pretty close between HP and Netgear - but HP pulled out all the stops and continued to slash their pricing so they came in about £1,500 cheaper than Netgear. It's not a huge amount of money in the grand scheme of things but when you consider the the HP kit has a much better feature set, a nice IOS (no horrid GUIs or web-based administration) and all of my cisco knowledge can pretty much be put to use straight away with HP it was a complete no-brainer.

Nice link from @Geoff there. Your Procurve kit should give you a pretty good idea of what's going on - check out the logs on the switches.

Sorry about the confusion... I'm not very good at multi-tasking without coffee!

Ok... I see... normal clients are nicking the addresses before you reserve them. So... do this: 1. Set up your pool as 10.xx.yy.10 --> 10.xx.yy.200. 2. Add an exclusion range of 10.xx.yy.10 --> 10.xx.yy.29 3. Activate your scope. 4. Add your reservations as and when you need to. That's it. Drink coffee, eat bacon and be happy.

^ What he said... Well worth spending a little bit of money on a Mac Mini server just for this. I did. It saved me hours. Plus it looks quite nice in my comms room.

Do you have a mac server?

You don't need to add an exclusion range - that would be used if you didn't use reservations but put static IPs on devices. Just set up your range, then put in all of your reservations and then hey-presto!

It depends on whether you want to explicitly set reservations for your printers in DHCP. If so - make the range 10.xx.yy.10 --> 10.xx.yy.200 - and then set reservations from .10 --> .29 for all your printers (you'll need to know the mac addresses though). Personally - I just give my printers fixed IPs (manually... not via DHCP) and use a different range for clients.

Thanks Synack - that second link is just what I was looking for!

I've recently noticed that there are two CA's on my network - one that my predecessor setup on a rather obscure server (which obviously was fantastically documented along the wealth of information that was left about everything else... ). Obviously I want to remove the older CA... but I notice that there are some certificates issued to my DCs using the "Domain Controller" certificate template. My question is this... what happens if I remove the old CA? Are these certificates of importance? or will my DCs just automatically sort themselves out? Or (!) will everything go horribly wrong? (Night becoming day, up is down, white is black, existence is nothingness... that sort of thing.) Cheers, Ant

One more thing - does your barracuda spam filter act as your receiving mail server? If so - does it authenticate with your Exchange 2010 box when it sends the mail on? By default the bog standard port 25 receive connector on Exchange 2010 requires authentication to keep it secure... Also - you could just allow anonymous users to connect to the Exchange 2010 receive connector...